WesternStates Coronavirus Survey€¦ · Fullsets of weighted frequencies and marginal percentages are available for each state in separate documents.The present document contains

Western States Coronavirus Survey Politics & Policy Results Montana State University‐Bozeman & University of Denver April 10‐27, 2020

Contents

Executive summary Pages 1‐3 Methods notes Page 4‐5 Colorado presidential results Pages 6‐9 Colorado Senate results Page 9‐12 Montana presidential results Pages 13‐16 Montana Senate results Pages 16‐19

Note: Full sets of weighted frequencies and marginal percentages are available for each state in separate documents. The present document contains presidential and Senate results only for the states of Colorado and Montana. However, the executive summary includes discussion of results for North Dakota and Utah, and the separate documents are available for these two states, as well.

Executive Summary

The Western States Coronavirus Survey was administered online between April 10 and April 27, 2020, with residents of Colorado (April 10‐19), Montana (April 10‐27), North Dakota (April 10‐25), and Utah (April 10‐15). A total of 2,220 individuals responded, with 503 in Colorado, 738 in Montana, 481 in North Dakota, and 498 in Utah. Margins of error associated with each state’s overall sample are reported in the methods notes section. That section also includes a description of the data weighting procedures. Analysis in this report is provided by Drs. Eric Raile, David Parker, and Elizabeth Shanahan of Montana State University‐Bozeman and by Dr. Pavielle Haines of the Center on American Politics at the University of Denver. This report focuses on public perceptions of responses to the novel coronavirus (COVID‐19) crisis, the status of Senate elections in Colorado and Montana, and the outlook for the fall presidential campaign.

Weighted marginal percentages and cross tabulations on pages 6‐19 provide certain state‐specific results. However, the following are key findings from the data across the four states:

Among political scientists, it is well known that the public rallies around their political leaders during times of crisis. We see this clearly at the state level. In every state, a large majority of people approve of the job their governor is doing to manage the coronavirus crisis. Although more people approve than disapprove of the job their U.S. senators are doing to manage the crisis (except in the case of Senator Cory Gardner in Colorado), 28‐40% of respondents report that they don’t know if they approve or disapprove. We suspect this is largely due to the volume of media attention concentrated on state and local officials. Relatively little coverage has been devoted to how specific senators or members of Congress are addressing the pandemic.

Report date: May 1, 2020 1

In Montana, North Dakota, and Utah, a majority of respondents approve of President Trump’s response to the coronavirus crisis. However, substantial minorities (about 40%) express disapproval. In Colorado, only 34% of voters approve of the job President Trump is doing related to the coronavirus crisis.

Respondents give overwhelmingly high marks to local governments, schools, scientists, and their fellow citizens for their response to the coronavirus crisis.

Both Montana and Colorado feature potentially competitive U.S. Senate races, with one acting Democratic governor (Montana) and one former Democratic governor (Colorado) challenging incumbent freshmen Republican senators. In Colorado, former governor John Hickenlooper has a substantial lead of 17 points that is outside the sample’s margin of error. But roughly 17% of likely Colorado voters remain undecided. Depending on which way they ultimately swing, Hickenlooper’s lead could shrink. Many undecided voters in Colorado are independents. In Montana –– despite having only announced his candidacy in early March –– Governor Steve Bullock is leading Senator Steve Daines 47‐40, with only 7% of likely voters still undecided. However, this lead is within the sample’s margin of error.

We also asked respondents in Montana and Colorado whom they would support in November’s presidential election. In Colorado, Democrat Joe Biden leads President Trump by 18 points (53% to 35%)—a substantial lead well outside the poll’s margin of error. Only 8% of voters are undecided, with another 3% expressing a preference for another candidate. In Montana, where Donald Trump bested Hillary Clinton by more than 20 points in 2016, we find a considerably tightened race. President Trump receives support from 45% of Montanans, while 40% of Montanans report they would vote for the Democratic nominee, Joe Biden. Trump’s lead is within the sample’s margin of error. Further, 10% indicate supporting some other candidate, with only 5% undecided.

Roughly 65‐80% of respondents in all four states believe state and local governments have had the appropriate level of involvement in developing policies in response to the coronavirus crisis. About 30% of respondents across three states—and 45% in Colorado—believe the federal government has not done enough in responding to the crisis. This breaks predictably along party lines, with Republicans generally believing the federal government has done about the right amount while Democrats are more likely to say not enough has been done. Independents generally fall somewhere in the middle, except in Colorado where a majority indicate that the federal government has not done enough in its coronavirus response.

People overwhelmingly support stay‐at‐home orders regardless of partisanship. However, some differences emerge when respondents are asked if people should be trusted to social distance on their own or if it is necessary for governments to issue orders because people cannot be trusted to voluntarily social distance themselves. Respondents are evenly split in Utah and Montana. A majority in Colorado indicate that governments must issue orders, while a majority in North Dakota trust people to take responsibility to socially distance themselves. In terms of partisan differences, Democrats overwhelmingly see a need for government orders in all four states, while a majority of Republicans support letting individuals take personal responsibility. Independents are split, with majorities in Montana and Colorado siding with government orders, while Independents in Utah


and North Dakota indicate a preference for trusting individuals to be responsible. These splits may hint at key differences in the political cultures of the four states.

Finally, the questionnaire asked whether the Chinese government shouldered any blame for the coronavirus crisis. Majorities in every state agree that the Chinese government is responsible because it withheld information about the pandemic. However, across every state Republicans and Independents are more likely to cast blame on China than are Democrats. Respondents were also asked whether they agreed or disagreed that it was unfair to blame China because the crisis could have erupted in any country. Pluralities in Montana and North Dakota disagree that it is unfair to blame China, while the split is roughly even in Colorado and Utah. These results suggest that the public is very sensitive to how elected officials choose to frame responsibility for the coronavirus crisis.


Methods Notes

The Western States Coronavirus Survey was administered online between April 10 and April 27, 2020, with residents of Colorado (April 10‐19), Montana (April 10‐27), North Dakota (April 10‐25), and Utah (April 10‐15). A total of 2,220 individuals responded, with 503 in Colorado, 738 in Montana, 481 in North Dakota, and 498 in Utah. The purpose of the survey was to assess public attitudes and self‐reported behaviors related to the coronavirus pandemic. A representative sample of respondents from each state was recruited from Qualtrics’s online probability panel. Except in special cases (see below), resulting data were weighted by age, gender, education, race, ethnicity, and urban versus rural residence to match U.S. Census Bureau data for each state. The survey was funded by Montana State University–Bozeman and the Center on American Politics at the University of Denver. The separate marginal percentage documents were produced by the Human Ecology Learning and Problem Solving (HELPS) Lab at Montana State University—Bozeman.

Sampling and Margin of Error Details

The population of interest included all adults in each target state. Based on the respective population size and number of respondents per state, the margins of error are:

+/‐ 4.4 percentage points for Colorado +/‐ 3.6 percentage points for Montana +/‐ 4.5 percentage points for North Dakota +/‐ 4.4 percentage points for Utah

These margins of error are the most conservative overall estimates for each state, using a confidence level of 95% and a response distribution of 50%. However, margins of error will be larger when analyzing subpopulations in cross tabular analyses. Respondents were screened out if they were under the age of eighteen, did not live in one of the target states, or finished the survey in less than one‐third of the median completion time.

Questionnaire

This 20‐minute online survey included a range of public health questions related to the coronavirus pandemic, as well as standard political and demographic items. Respondents were asked about their level of concern surrounding the coronavirus pandemic, the level of disruption to their lives, their social distancing practices, and their preferred policy solutions, among other topics. When possible, language was modeled after items from the American National Election Study, the Pew Research Center, Gallup, and General Social Survey.

Weighting Procedure

To ensure that the sample of respondents reflected the adult population in each state, the samples were weighted using iterative proportional fitting. Also known as raking, this procedure forces sample margins to approximate population margins for key demographic traits. These traits included age, gender, education, race, ethnicity, and urban versus rural residence. Based on these characteristics, a weight was assigned to each respondent. Respondents from under‐sampled demographic groups received larger weights to ensure their attitudes and behaviors were fairly represented in the aggregate results. For instance, the CDC reported that 48.1% of Coloradans received a flu shot during the 2018‐2019 flu season. In the unweighted Colorado sample, 58.9% of respondents said they got a flu shot. After weighting, a


more accurate 48.9% of the sample said they got a flu shot. Weights were truncated at 6.0 to avoid overweighting the data.

However, the weighting procedure for the presidential and Senate election questions for the states of Colorado and Montana was different. The raking procedure described above was applied to a subsample of individuals who are registered to vote, are following campaigns, and indicate they are likely to vote. The variables used in the weighting procedure were respondent age, gender, education level, race, and 2016 presidential vote.


Colorado Weighted Results

Notes: The tables below show marginal percentages and cross tabulations of variables. Row percentages may not equal 100% exactly due to rounding. The information in parentheses for each row (n=) is the number of observations for that category overall. The data are weighted to the population of registered voters in the state after screening for likely voters as described in the methods notes. Partisan response options were randomized.

Table 1: Colorado Intended Presidential Vote

Response Categories (n=379) Donald Trump, Joe Biden, Republican Democrat Other Don’t know

35% 53% 3% 8%

Question text: If the general election for U.S. President were held today, for whom would you vote?

Brief discussion: Biden holds a considerable 18‐point lead over Trump in Colorado. This is outside the margin of error and only 8% of likely voters remain undecided, suggesting that Biden is likely to win.

Table 2: Colorado Intended Presidential Vote by Partisan Identification Donald Trump, Joe Biden,

Response categories Republican Democrat Other Don’t know

Democrat (n=141) 4% 89% 2% 4%

Republican (n=95) 88% 8% 1% 3%

Independent (n=137) 30% 47% 6% 16%

Something else (n=6) 15% 67% 3% 15%


Brief discussion: Colorado Democrats and Republicans overwhelmingly intend to vote along party lines. Independents and those who identify with a different party favor Biden, but a large percentage remain undecided.


Table 3: Colorado Intended Presidential Vote by Political Ideology Donald Trump, Joe Biden,


Extremely liberal (n=39) 0% 96% 4% 0%

Liberal (n=44) 0% 96% 3% 2%

Slightly liberal (n=48) 0% 94% 3% 4%

Moderate (n=129) 37% 44% 5% 13%

Slightly conservative (n=46) 60% 32% 2% 6%

Conservative (n=49) 71% 8% 2% 19%

Extremely conservative (n=23) 98% 1% 0% 1%


Brief discussion: Liberal Coloradans are strongly committed to voting for Biden. Moderates are more divided, but also tend to indicate greater support for Biden. Although a majority of conservative Coloradans intend to vote for Trump, a noteworthy minority intend to vote for Biden or remain undecided.

Table 4: Colorado Intended Presidential Vote by Gender Donald Trump, Joe Biden,


Female (n=186) 37% 50% 5% 8%

Male (n=190) 32% 57% 2% 9%


Brief discussion: Men and women in Colorado have similar vote intentions for the 2020 presidential election. This is unusual, as there is generally a large gender gap, with women being more supportive of the Democratic candidate.


Table 5: Colorado Intended Presidential Vote by Age Donald Trump, Joe Biden,


Ages 18‐29 (n=61) 33% 52% 0% 14%

Ages 30‐39 (n=102) 31% 54% 2% 14%

Ages 40‐49 (n=68) 29% 66% 2% 2%

Ages 50‐59 (n=58) 35% 49% 11% 4%

Ages 60‐69 (n=67) 42% 47% 4% 6%

Ages 70+ (n=23) 55% 41% 1% 3%


Brief discussion: Younger Colorado voters tend to be less certain about their vote choice, but a majority indicate a preference for Biden. Middle‐aged voters are more certain of their choice, and at least a plurality say they will vote for Biden. Voters 70 and older swing heavily for Trump.

Table 6: Colorado Intended Presidential Vote by Household Income Donald Trump, Joe Biden,


$30,000 or less (n=122) 26% 58% 3% 13%

$30,001 ‐ $50,000 (n=50) 49% 28% 5% 17%

$50,001 ‐ $100,000 (n=104) 38% 56% 2% 3%

$100,001 ‐ $150,000 (n=65) 38% 59% 2% 2%

More than $150,000 (n=37) 33% 55% 6% 6%


Brief discussion: Across nearly every income bracket, Biden has a sizable advantage over Trump. The one exception is for voters making $30‐50K a year, who swing heavily for Trump.


Table 7: Colorado Intended Presidential Vote by Education Level Donald Trump, Joe Biden,


Less than high school (n=6) 100% 0% 0% 0%

HS diploma / GED (n=166) 33% 50% 4% 13%

Associate degree (n=53) 56% 35% 1% 9%

Bachelor’s degree (n=98) 32% 61% 3% 4%

Graduate degree (n=56) 20% 71% 5% 4%


Brief discussion: Coloradans with at least a bachelor’s degree swing heavily for Biden. Those with less education are more split between Trump and Biden, and many remain uncertain of their vote choice.

Table 8: Colorado Intended Senate Vote

Response Categories (n=379) John

Hickenlooper, Cory Gardner, Democrat Republican Other Don’t know

48% 31% 4% 17%

Question text: If the general election for U.S. Senate were held today, for whom would you vote?

Brief discussion: Hickenlooper holds a considerable 17‐point lead over Gardner. This is outside the margin of error, but 17% of likely voters remain undecided. This means that although a Hickenlooper victory is likely, the race could become much tighter depending on whom undecided voters ultimately decide to support.


Table 9: Colorado Intended Senate Vote by Partisan Identification John

Hickenlooper, Cory Gardner, Response categories Democrat Republican Other Don’t know

Democrat (n=141) 85% 1% 2% 12%

Republican (n=95) 6% 81% 1% 11%

Independent (n=137) 41% 28% 7% 24%



Brief discussion: Colorado Democrats and Republicans break along party lines. While a large portion of independents remain undecided, a plurality support Hickenlooper.

Table 10: Colorado Intended Senate Vote by Political Ideology John



Liberal (n=44) 94% 2% 1% 3%


Moderate (n=129) 39% 28% 7% 27%


Conservative (n=49) 5% 60% 3% 32%



Brief discussion: Liberal Coloradans support Hickenlooper, while conservative Coloradans support Gardner. Although a large portion of moderates remain undecided, a plurality support Hickenlooper.


Table 11: Colorado Intended Senate Vote by Gender

Response categories

John Hickenlooper, Democrat

Cory Gardner, Republican Other Don’t know

Female (n=186) 43% 34% 3% 21%

Male (n=190) 53% 28% 4% 14%


Brief discussion: Men and women are both more likely to vote for Hickenlooper than for Gardner.

Table 12: Colorado Intended Senate Vote by Age John


Ages 18‐29 (n=61) 49% 10% 7% 34%

Ages 30‐39 (n=102) 41% 36% 4% 19%

Ages 40‐49 (n=68) 63% 29% 1% 8%

Ages 50‐59 (n=58) 50% 35% 2% 13%

Ages 60‐69 (n=67) 44% 43% 5% 8%

Ages 70+ (n=23) 33% 29% 7% 31%


Brief discussion: Younger Colorado voters tend to be less certain of their vote choice, but a plurality indicate Hickenlooper. Middle‐aged voters are more certain of their choice, and a majority say they will vote for Hickenlooper. Older voters ages 60‐69 and 70+ are evenly split between Hickenlooper and Gardner. This signals trouble for Gardner, since older voters are usually a reliable Republican base.


Table 13: Colorado Intended Senate Vote by Household Income John


$30,000 or less (n=122) 54% 19% 5% 23%

$30,001 ‐ $50,000 (n=50) 28% 48% 5% 19%

$50,001 ‐ $100,000 (n=104) 49% 41% 4% 6%

$100,001 ‐ $150,000 (n=65) 54% 25% 2% 19%

More than $150,000 (n=37) 41% 32% 1% 27%


Brief discussion: Across nearly every income bracket, Hickenlooper has an advantage over Gardner. The one exception is for voters making $30‐50K a year, who prefer Gardner.

Table 14: Colorado Intended Senate Vote by Education Level John



HS diploma / GED (n=166) 47% 31% 4% 18%





Brief discussion: Coloradans with at least a bachelor’s degree swing heavily for Hickenlooper. Those with less education are more split between Hickenlooper and Gardner, and many remain uncertain of their vote choice.


Montana Weighted Results

Notes: The tables below show marginal percentages and cross tabulations of variables. Row percentages may not equal 100% exactly due to rounding. The information in parentheses for each row (n=) is the number of observations for that category overall. The data are weighted to the population of registered voters in the state after screening for likely voters as described in the methods notes. Partisan response options were randomized.

Table 1: Montana Intended Presidential Vote

Response Categories (n=459) Donald Trump, Joe Biden, Republican Democrat Other Don’t know

45% 40% 10% 5%


Brief discussion: The margin of error for this question is +/‐ 4.6 percentage points, so the presidential results are within the sample’s margin of error. Note that Biden outperforms Hillary Clinton by roughly 6 percentage points and is matching President Obama’s 2012 vote share. President Trump is ten points below his 2016 performance.

Table 2: Montana Intended Presidential Vote by Partisan Identification Donald Trump, Joe Biden,


Democrat (n=155) 2% 90% 7% 1%

Republican (n=170) 93% 0% 5% 1%

Independent (n=112) 33% 34% 17% 16%

Something else (n=22) 35% 17% 44% 4%


Brief discussion: In an age of strong partisanship, Republicans and Democrats are staying put with the presumptive nominees of their respective parties. Independents make up an important segment of the Montana electorate and lean conservative. Although independents swung heavily for Democrat Jon Tester in the 2018 Senate race, they are split evenly between President Trump and former Vice President Biden with considerable uncertainty. The margin among independents and the fact that so many are uncommitted currently is perhaps suggestive of a closer presidential contest in 2020 than 2016, which is consistent with the finding that the presidential race is within the poll’s margin of error.


Table 3: Montana Intended Presidential Vote by Political Ideology Donald Trump, Joe Biden,



Liberal (n=76) 6% 67% 25% 3%


Moderate (n=114) 28% 50% 7% 15%


Conservative (n=91) 96% 3% 1% 1%



Brief discussion: There is some evidence of softening support for President Trump among conservative voters. Although the president dominates among self‐described conservatives, 21 percent of extremely conservative voters indicate a preference for Biden or some other candidate—and 11 percent of slightly conservative voters support someone other than Trump. Moderates also heavily prefer Biden. In the 2018 election, conservative and moderate defections to Democrat Senator Jon Tester helped him win reelection narrowly.

Table 4: Montana Intended Presidential Vote by Gender Donald Trump, Joe Biden,


Female (n=236) 39% 44% 11% 6%

Male (n=218) 54% 34% 9% 3%


Brief discussion: Men heavily favor President Trump, while women lean slightly toward Joe Biden. Interestingly, the gender gap for President Trump is similar to that faced by Republican Matt Rosendale in the 2018 Senate race. If undecided women break to Biden, this would tighten the presidential race even more.


Table 5: Montana Intended Presidential Vote by Age Donald Trump, Joe Biden,


Ages 18‐29 (n=131) 24% 47% 21% 7%

Ages 30‐39 (n=120) 51% 36% 7% 5%

Ages 40‐49 (n=82) 59% 28% 9% 4%

Ages 50‐59 (n=57) 51% 44% 1% 4%

Ages 60‐69 (n=43) 49% 49% 1% 1%

Ages 70+ (n=26) 61% 27% 10% 2%


Brief discussion: Only the youngest cohort of voters supports Joe Biden, but Biden’s relative strength among voters aged 60‐69 should give Republicans some concern. Democrat Jon Tester won these voters in 2018.

Table 6: Montana Intended Presidential Vote by Household Income Donald Trump, Joe Biden,


$30,000 or less (n=179) 35% 40% 20% 5%

$30,001 ‐ $50,000 (n=74) 63% 31% 3% 3%

$50,001 ‐ $100,000 (n=136) 47% 40% 7% 6%

$100,001 ‐ $150,000 (n=44) 46% 52% 0% 2%

More than $150,000 (n=26) 60% 37% 0% 2%


Brief discussion: The wealthiest voters support the reelection of President Trump, while the lowest‐income voters have a slight preference for Biden. Middle‐class voters appear more evenly split between Biden and President Trump.


Table 7: Montana Intended Presidential Vote by Education Level Donald Trump, Joe Biden,



HS diploma / GED (n=257) 49% 34% 16% 1%





Brief discussion: The education divide that emerged in the 2016 presidential election nationally is replicated here. Montanans with a four‐year degree or post‐graduate education strongly support Joe Biden. Voters with two‐year degrees, high school diplomas, or less, however, support the reelection of President Trump. The high percentage of “don’t know” responses among those with less than a high school degree is an artifact of the small sample size.

Table 8: Montana Intended Senate Vote

Response Categories (n=458) Steve Daines, Steve Bullock, Republican Democrat Other Don’t know

40% 47% 6% 7%


Brief discussion: The overall margin of error for this question is +/‐ 4.6 percentage points, so the Senate race results are within the margin of error.


Table 9: Montana Intended Senate Vote by Partisan Identification Steve Daines, Steve Bullock,


Democrat (n=156) 3% 87% 6% 3%

Republican (n=172) 82% 9% 6% 3%

Independent (n=107) 32% 46% 7% 15%



Brief discussion: There are three key takeaways here. First, given the polarization of the political environment, Republican Senator Steve Daines should be concerned that 9 percent of Republicans indicate a preference for Democrat Steve Bullock while only 3 percent of Democrats indicate a willingness to vote for Daines. Similar levels of Republican defection in 2018 and 2012 were key to Democratic Senator Jon Tester’s reelection in both contests. Second, for Democrats to win, they need independent voters— and right now, Bullock leads among those voters. Finally, Bullock enjoys less support among Democrats than Democratic Senator Jon Tester did in 2018. Tester received near universal support from Democrats casting ballots in that election.

Table 10: Montana Intended Senate Vote by Political Ideology Steve Daines, Steve Bullock,



Liberal (n=77) 2% 89% 5% 4%


Moderate (n=115) 26% 57% 1% 15%


Conservative (n=92) 80% 8% 8% 4%



Brief discussion: In 2018, Republican Senate candidate Matt Rosendale dominated among extremely conservative and conservative voters, winning 93 and 95 percent of them, respectively. The finding that only 72 percent and 80 percent of extremely conservative and conservative voters indicate they will vote for Senator Daines should be of serious concern to him and his campaign. Liberal voters, by comparison, are more strongly committed for Steve Bullock, though less so with extremely liberal voters. Given that the Green Party and Libertarian Party will have candidates on the ballot this fall, it is possible that both major‐party candidates could lose support from extremely ideological voters as they gravitate toward these parties.


Table 11: Montana Intended Senate Vote by Gender Steve Daines, Steve Bullock,


Female (n=239) 29% 52% 6% 13%

Male (n=214) 53% 40% 6% 1%


Brief discussion: In the 2018 Montana Senate race, the gender gap for Republican Senate candidate Matt Rosendale was 10 points. Senator Steve Daines is facing a gap of more than twice that—while Democrat Steve Bullock’s support among men and women is nearly identical to Senator Jon Tester’s in 2018. This suggests that perhaps Republican women are open to defecting to Steve Bullock and may help explain why our poll shows a statistically tied race. Given that women are far more likely to express stress and worry about the coronavirus, they may also be more willing to support a governor who has been dominating media coverage as someone working to address the crisis.

Table 12: Montana Intended Senate Vote by Age Steve Daines, Steve Bullock,


Ages 18‐29 (n=127) 29% 47% 11% 13%

Ages 30‐39 (n=121) 46% 45% 3% 7%

Ages 40‐49 (n=83) 44% 47% 5% 4%

Ages 50‐59 (n=58) 33% 53% 7% 7%

Ages 60‐69 (n=43) 45% 52% 0% 3%

Ages 70+ (n=27) 57% 33% 9% 0%


Brief discussion: There is widespread support for Democrat Steve Bullock across age groups, with only the very oldest voters expressing solid support for Senator Daines. Older voters are key for Republican electoral success, and Senator Daines will need to narrow the gap among voters between 50 and 69 years of age in order to win.


Table 13: Montana Intended Senate Vote by Household Income Steve Daines, Steve Bullock,


$30,000 or less (n=175) 36% 44% 11% 9%

$30,001 ‐ $50,000 (n=75) 46% 40% 5% 9%

$50,001 ‐ $100,000 (n=138) 38% 53% 3% 6%

$100,001 ‐ $150,000 (n=45) 37% 54% 2% 7%

More than $150,000 (n=26) 62% 38% 0% 0%


Brief discussion: The wealthiest Montanans overwhelmingly support the reelection of Senator Daines. However, Montana is not a wealthy state. Bullock has majority support from every other income group except for respondents making between $30,000 and $50,000.

Table 14: Montana Intended Senate Vote by Education Level Steve Daines, Steve Bullock,



HS diploma / GED (n=254) 45% 41% 8% 6%





Brief discussion: The education divide, apparent in the presidential race, is also apparent here. Voters with any education beyond high school support Bullock, while those with a high school degree narrowly support Daines’ reelection.


Documents

WesternStates Coronavirus Survey€¦ · Fullsets of weighted frequencies and marginal percentages are available for each state in separate documents.The present document contains